1988 年 19 巻 p. 257-274
The central theme of this paper concerns the spatial dichotomy between innovative firms in the manufacturing sector. In this respect we will concentrate on the question whether and to which degree relatively innovative (especially small) industrial firms are spatially concentrated, and more in particular whether they are located in areas with a favourable selection environment. This issue is treated here from the viewpoint of the validity of the so-called urban symbiosis hypothesis and its complementary selective centrifugality hypothesis.
For this purpose we constructed three latent variables, viz. the concepts ‘selection environment’, ‘innovation potential’ and ‘innovativeness’ whose interrelationships will be estimated-on the basis of multiple indicators-by means of the Partial Least Squares method.
The first concept is related to the locational profile of the region in which the firm is located, while the latter two concepts refer to the intra-firm characteristics (number of R&D employees, number of innovations and so on), and more precisely to innovation input and innovation output indicators respectively. In analysing the relations between these three concepts our main research finding is that a great many (urban) regions in the central parts of the Netherlands (the so-called Rimcity or Randstad) appear to be in a rather unfavourable position regarding innovative capacity of small industrial firms, even for firms which on average are assumed to have favourable technological prospects. The results cast some doubt on the validity of the urban symbiosis hypothesis, whereas the selective centrifugality hypothesis appears to be more valid in the Dutch context.